Skip to content

Support lookup custom shuffle by Fluss bucket based on FLIP-462 #3727

Description

@loserwang1024

Search before asking

  • I searched in the issues and found nothing similar.

Motivation

Currently, Fluss lookup join does not support lookup custom shuffle. When lookup join is executed with normal hash shuffle, records are distributed by Flink's join-key hash, which is not aligned with Fluss bucket assignment.

This makes it difficult to combine lookup join with full-cache lookup efficiently. Each lookup subtask may receive keys from almost all Fluss buckets, so every subtask may need to load a large portion of the dimension table.

This issue aims to support lookup custom shuffle by Fluss bucket in Flink 2.2, so lookup input records can be shuffled according to Fluss bucket rules.

Solution

No response

Anything else?

No response

Willingness to contribute

  • I'm willing to submit a PR!

Metadata

Metadata

Assignees

Labels

No labels
No labels

Fields

No fields configured for Feature.

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions